Health and Safety Executive 2008 Evaluating Protection Aforded by Surgical Masks

Health and Safety
Executive
Evaluating the protection afforded by surgical

masks against influenza bioaerosols
Gross protection of surgical masks compared to
filtering facepiece respirators
Prepared by the Health and Safety Laboratory

for the Health and Safety Executive 2008
RR619
Research Report
Health and Safety
Executive

Jonathan Gawn, Mike Clayton,

Catherine Makison & Brian Crook
Health Improvement and Human Factors Groups
Health and Safety Laboratory
Harpur Hill
Buxton
SK17 9JN
The UK is preparing for a potential influenza pandemic. The main route of transmission of influenza is believed to be
via direct contact with large droplets. The relative importance of aerosols in transmission is considered to be minor, but
it cannot be ruled-out. The current UK Pandemic Influenza Infection Control Guidance recommends that workers who
are in close contact with patients should wear surgical masks to reduce exposure to large droplets. However, surgical
masks are not intended to provide protection against infectious aerosols. The guidance recommends that procedures that
are likely to generate aerosols should be minimised, or where unavoidable, workers should wear appropriate respiratory
protection. There is a common misperception amongst workers and employers that surgical masks will protect against
aerosols. This study aims to evaluate the relative levels of protection provided by both surgical masks and respirators
against aerosols.
This study focussed on the effectiveness of surgical masks against a range of airborne particles. Using separate tests to
measure levels of inert particles and live aerosolised influenza virus, our findings show that surgical masks provide around
a 6-fold reduction in exposure. Live viruses could be detected in the air behind all surgical masks tested. By contrast,
properly fitted respirators could provide at least a 100-fold reduction.
This report and the work it describes were funded by the Health and Safety Executive (HSE). Its contents, including any
opinions and/or conclusions expressed, do not necessarily reflect HSE policy.
HSE Books
© Crown copyright 2008
First published 2008
All rights reserved. No part of this publication may be
reproduced, stored in a retrieval system, or transmitted
in any form or by any means (electronic, mechanical,
photocopying, recording or otherwise) without the prior
written permission of the copyright owner.
Applications for reproduction should be made in writing to:
Licensing Division, Her Majesty’s Stationery Office,
St Clements House, 2-16 Colegate, Norwich NR3 1BQ
or by e-mail to hmsolicensing@cabinet-office.x.gsi.gov.uk
ii
SUMMARY
Background
The UK is preparing for a potential influenza pandemic. The main route of transmission of
influenza is believed to be via direct contact with large droplets. The relative importance of
aerosols in transmission is considered to be minor, but it cannot be ruled-out. The current UK
Pandemic Influenza Infection Control Guidance recommends that workers who are in close
contact with patients should wear surgical masks to reduce exposure to large droplets.
However, surgical masks are not intended to provide protection against infectious aerosols. The
guidance recommends that procedures that are likely to generate aerosols should be minimised,
or where unavoidable, workers should wear appropriate respiratory protection. There is a
common misperception amongst workers and employers that surgical masks will protect against
aerosols. This study aims to evaluate the relative levels of protection provided by both surgical
masks and respirators against aerosols.
Main Findings
This study focussed on the effectiveness of surgical masks against a range of airborne particles.
Using separate tests to measure levels of inert particles and live aerosolised influenza virus, our
findings show that surgical masks provide around a 6-fold reduction in exposure. Live viruses
could be detected in the air behind all surgical masks tested. By contrast, properly fitted
respirators could provide at least a 100-fold reduction.
iii
iv
CONTENTS
EXECUTIVE SUMMARY ................................................................................. VII
1 INTRODUCTION......................................................................................1
1.1 Background ....................................................................................................... 1
1.2 Respiratory Protective Equipment ..................................................................... 2
1.3 RPE in Healthcare and Pandemic Planning ...................................................... 3
2 MATERIALS AND METHODS.................................................................7
2.1 Inert Particle Tests On Test Subject .................................................................. 7
2.2 Respirators And Surgical Masks Employed In The Study .................................. 9
2.3 Influenza Bioaerosol Tests .............................................................................. 10
3 RESULTS...............................................................................................13
3.1 Inert Particle Exposure Tests On Test Subject ................................................ 13
3.2 Influenza Bioaerosol Tests .............................................................................. 18
4 DISCUSSION .........................................................................................20
4.1 Fitted Facepiece Respirators........................................................................... 20
4.2 Surgical Masks................................................................................................ 20
5 CONCLUSIONS .....................................................................................24
6 RECOMMENDATIONS FOR FUTURE WORK......................................25
6.1 Testing A Broader Range Of Surgical Masks And Respirators ........................ 25
6.2 Maximising The Protective Effect of Surgical Masks Against Influenza
Bioaerosols ..................................................................................................... 25
6.3 Analysis Of Surgical Mask Protection Against A More Representative
Influenza Bioaerosol........................................................................................ 25
6.4 Testing Of The Survival And Dissemination Of Live Influenza Beyond One
Metre............................................................................................................... 26
6.5 Bioaerosol Challenge To FFP Respirators ...................................................... 26
7 APPENDICES ........................................................................................27
7.1 Details Of Products Available Via NHS Logistics (List Received At HSL
Feb May 2006) ................................................................................................ 27
7.2 Details Of Products Available Via NHS Logistics (List Received At HSL
May 2006) ....................................................................................................... 28
8 REFERENCES .......................................................................................29
9 GLOSSARY ...........................................................................................33
v
vi
EXECUTIVE SUMMARY
Objectives
The UK is advanced in its preparations for a potential outbreak of human pandemic influenza.
The main route of transmission of influenza is believed to be via large droplets or direct contact
with secretions and, in some circumstances, exposure to infectious aerosols.
The relative contribution of aerosol transmission in natural influenza transmission is thought to

be minor but cannot be ruled-out. The likelihood of infection via this route will increase when
in close proximity to the patient and especially when carrying out procedures likely to generate
aerosols, such as intubation or dental drilling. Consequently, the current UK Pandemic
Influenza Infection Control Guidance recommends the wearing of fluid-repellent surgical masks
for those workers who are in close contact with symptomatic patients as protection from
droplets/splashes and recommends the use of respiratory protection (i.e. FFP3 respirators) for
circumstances in which aerosols are generated as a consequence of medical procedures.
Whilst surgical masks may, in principle, offer adequate protection against large droplets and
contact transmission, the level of protection they offer against a residual aerosol risk is poorly
understood. They are not designed, or certified, as respiratory protective devices. However,
there is a common misperception that they will provide protection against aerosols.
The following study aims to measure the efficiency of surgical masks against airborne particles
generated from a simulated sneeze (including those that contain live, infectious influenza virus)
so that the contribution of surgical masks in the protection against any residual aerosol risk can
be assessed.
Main Findings
Surgical masks and filtering facepiece (FFP) respirators were tested on a human volunteer using
an inert aerosol challenge. From the results of this study, it can be concluded that surgical
masks will mitigate a mean reduction factor of around 2 against a simulated sneeze of inert
airborne particles compared to FFP respirators, which are capable of offering a mean reduction
factor of 100 or higher.
Surgical masks were also tested on a breathing dummy head and subjected to an aerosol
challenge containing live influenza virus. Infectious, viable virus could be detected in the air
behind all surgical masks challenged. A mean reduction factor of 6 was measured.
Recommendations
In principle, surgical masks that are worn correctly should provide adequate protection against
large droplets, splashes and contact transmission. They may also reduce to some degree any
residual aerosol risk, although this level of protection might not sufficiently reduce the
likelihood of transmission via this route. Consequently they should not be used in situations
where close exposure to infectious aerosols is likely.
With this in mind, it is recommended that HSE draw the results of this research to the attention
of DH/HPA so that it can be considered as part of the wider issue of modes of influenza
transmission.
vii
viii
1 INTRODUCTION
1.1 BACKGROUND
There is growing concern that an H5N1 avian influenza virus currently causing global outbreaks
of disease in wild and domestic poultry could be a candidate virus for a human pandemic. This
is partially due to the ability of this virus to cause serious and often fatal diseases in humans
who have been in close contact with infected birds. Although no human-to-human transmission
of the H5N1 virus has yet been observed, its ability to replicate in humans might permit
adaptation, resulting in an antigenically novel virus that could cause a global epidemic, excess
hospitalisations and fatalities. In any event, pandemics occur sporadically and even if the
current H5N1 candidate fails to adapt to humans, it is widely expected that another strain of the
virus will do so in due course.
Health protection organisations worldwide are making preparations for such an outbreak and
national risk assessments and pandemic contingency plans are at an advanced stage in the UK
(HPA 2005; DH 2007a; NHS 2007). Many of these plans offer advice to healthcare workers
regarding precautions to take in order to avoid exposure to infectious aerosols generated by
people suffering from pandemic influenza, or from certain medical procedures carried out on
them. Clearly, there is a need for front-line healthcare workers to be adequately protected in
the event of such an outbreak, which will require organisational, environmental and procedural
controls and may rely heavily on effective protective equipment (Yassi et al., 2005).
In theory, influenza can be transmitted via three major routes
• Large droplets (>5µm diameter)– these settle rapidly on surfaces
• Manual inoculation/direct contact or contact with secretions
• Aerosols – small droplets (<5µm diameter) that remain airborne for protracted periods
Prevailing expert opinion is that the balance of evidence points to large droplet and
direct/indirect contact as being the most important routes of influenza transmission. The relative
contribution of aerosols in natural transmission of the virus is unknown, but is believed to be
minor (Brankston et al., 2007; Bell, 2006; Bridges et al., 2003; Gardam and Lemeiux 2007; Lee
2007; DH 2007b). Consequently, many pandemic planning bodies, including those in the UK,
have placed less importance on infectious aerosols as being a natural feature of influenza
transmission and indicate personal protection against bioaerosols only for certain aerosol
generating procedures (HPA 2005; DH 2007a; NHS 2007). This view of transmission is based
upon the rarity of long-range infections, which suggests that the ability of the virus to remain
viable in aerosols is limited. However, the rarity of long-range infections does not preclude a
role for bioaerosols in the transmission of influenza virus in close proximity. The evidence for
an aerosol component to natural transmission of influenza virus was recently reviewed and it
was concluded that they may have a role (Tellier, 2006; Toner, 2006; see also Tellier 2007;
Tang and Li 2007). Influenza can be transmitted to both humans and mice via artificially
generated aerosols (Alford et al., 1966; Loosli et al., 1943; Schulman, 1967) and via the air
between mice and birds in a laboratory setting (Schulman, 1968; Webster et al., 2002).
Therefore, the three modes of transmission are not mutually exclusive and it is difficult to
separate them in order to draw conclusions as to the relative importance of each. Whilst the
absence of long-range infections is evidential for large droplet and contact transmission in a
common infection, it remains that it is reasonable to consider that multiple routes of
transmission may be possible and that infectious aerosols, regardless of their relative importance
in the context of other modes, should not be discounted.
1
Given the potential severity of the consequences associated with contracting a pandemic strain
of influenza, it is clear that exposure of healthcare workers and others that are to work in close
proximity to infected individuals should be minimised, that includes the selection of adequate
protective equipment that is suitable for the tasks required.
1.2 RESPIRATORY PROTECTIVE EQUIPMENT
The Control of Substances Hazardous to Health Regulations 2002 (COSHH) covers not only
exposure to hazardous chemicals but also biological agents. The regulations assert that ‘every
employer shall ensure that the exposure of his employees to substances hazardous to health is
either prevented or, where this is not reasonably practicable, adequately controlled’. HSE
recommends a hierarchy of principles for the control of exposure to airborne hazardous
substances to underpin this requirement. These are:
• Elimination
• Substitution
• Physical separation
• Use of personal protective equipment (PPE)
Vaccination can be added to this list of control measures but should not be seen as a primary
measure as it might preclude the principles of controlling exposure to an organism.
Where PPE is employed as a control measure then it should be ‘adequate’ for the anticipated
exposure levels and ‘suitable’ for the task, for the environment and for the wearer. In addition
the PPE must be:
• “CE” marked to the European PPE Directive
• Selected, used and maintained by properly trained people
• Correctly maintained, examined and tested
• Correctly stored
The British Standard BS EN 149:2001 covers disposable filtering facepiece (FFP) respirators.
FFP respirators are classified as FFP1, FFP2 and FFP3 according to the level of protection
afforded as assessed by specified laboratory tests, with FFP3 offering the most protection. In
order to aid the correct selection of ‘adequate’ RPE assigned protection factors (APF) have
been derived, and for FFP respirators these are 4, 10 and 20 respectively. The APF is the ratio
of pollutant outside the device to that inside the device and is defined by British Standard BS
EN 529:2005 as the ‘level of respiratory protection that can realistically be expected to be
achieved in the workplace by 95% of adequately trained and supervised wearers using a
properly functioning and correctly fitted respiratory protective device and is based on the 5th
percentile of the Workplace Protection Factor (WPF) data’. APFs are published by both BS
EN 529:2005 and by HSE in its RPE guidance HSG53 (HSE 2005a). Table 1 shows the
efficiency requirement levels for the three classes of filtering facepieces from British Standards,
together with their assigned protection factors.
Disposable filtering facepieces can only be said to be suitable for the wearer if they have been
subject to, and have passed, a ‘Fit Test’ on the intended wearer. A fit test is a means of
2
assessing the goodness of fit of the facepiece to the wearer’s face. HSE Operational Circular
OC282/28 provides further information (HSE 2004).
The European PPE Directive 89/686/EEC covers Respiratory Protective Equipment. This
Directive excludes surgical masks and they are not certified for use as RPE in the UK. Surgical
masks can be certified compliant with the Medical Devices Directive and be ‘CE’ marked.
However, the placing of a ‘CE’ mark on a surgical mask does not denote the ability to provide
respiratory protection under the PPE Directive. Whilst surgical masks do provide a degree of
protection against droplets and splashing, the British Standard covering surgical masks (BS EN
14683:2005) categorically states that ‘The surgical masks intended to be used in operating
theatres and health care settings with similar requirements are designed to protect the working
environment and not the wearer. When the primary intention is to protect the wearer from
infection, the use of respiratory protective devices should be considered’. As surgical masks are
not intended to offer protection against airborne particles, they are not designed to fit closely to
the wearers face or designed to have the filtering efficiencies required for adequate respiratory
protection. Furthermore, no protection factors are assigned to surgical masks, as they are not
designed to offer respiratory protection. However, there is a common misperception that they
will provide protection against aerosols.
Table 1. Efficiency requirement for filtering facepieces and their assigned

protection factors
Max Max
Min Nominal Assigned
permitted permitted
Class filter Protection Protection
total inward filter
efficiency� Factor   Factor
leakage penetration
FFP1 22% 20 80% 4.5 4
FFP2 8% 6 94% 12.5 10
FFP3 2% 1 99% 50 20
� Figure derived from the maximum filter penetration allowed by BS EN 149:2001
�� Figure derived from the maximum total inward leakage allowed by BS EN 149:2001
1.3 RPE IN HEALTHCARE AND PANDEMIC PLANNING
Healthcare workers are exposed to infectious agents as a consequence of their work. In the
context of the COSHH Regulations, elimination, substitution, and physical separation are not
possible in the healthcare setting. Since physically preventing exposure of healthcare workers
to the virus is not feasible, it is important to minimise the likelihood that they will become
infected, as far as is reasonably practicable, whilst still ensuring they are able to undertake their
duties effectively. What is both reasonable and practicable will change during a pandemic,
although the duty of control will still be based upon applying protective measures appropriate to
the activity and consistent with the risk assessment. The UK Pandemic Influenza Infection
Control Guidance identifies approaches, systems of work and infection control measures to
protect healthcare workers. This places a heavy reliance on the use of personal protective
equipment (including RPE) and vaccination. In the first waves of a pandemic, it is unlikely that
a protective vaccine will be available. Therefore, PPE (including RPE) will play a major part in
the control programme. As a result, it is important to be able to correctly select the appropriate
type and class of PPE and RPE and to be aware of any limitations that may exist in the
protection it affords.
The type of RPE that is considered to be most suitable for use by healthcare workers are FFP
respirators. The use of FFP respirators would necessitate correct maintenance, correct storage,
3
fit testing and use by trained personnel. These aspects, as well as respirator supply, would
clearly present a planning challenge to the healthcare sector.
The current UK Pandemic Influenza Infection Control Guidance (HPA 2005; NHS 2007), and
the draft framework for responding to an influenza pandemic (DH 2007a) recommends the
wearing of class FFP3 disposable respirators when carrying out clinical procedures likely to
generate aerosols of respiratory secretions from infected patients (e.g. dental drilling,
intubations, aspiration). For those workers who may be in close or frequent contact with
symptomatic patients (i.e. within one metre), the use of fluid-repellent surgical masks is
recommended as protection from large droplets or splashes. Therefore, the guidance recognises
there may be a hazard from infectious aerosols from certain procedures and prescribes suitable
control measures. However, there is concern that natural processes such as coughing, sneezing
and talking may also generate infectious aerosols resulting in a residual risk of infection.
Whether this is sufficient to cause an infection is unknown.
The scant data regarding the distribution of particle sizes in coughs and sneezes was recently
reviewed by Nicas et al (Nicas et al., 2005) and analysed in relation to the likelihood of
infection via the inhalation route. A single sneeze contains more particles than a single cough,
but particles in both a sneeze and a cough are thought to vary in size from <1 to >2000µm, with
the large majority being in the <20µm range. These could be inhaled and therefore pose a risk
of infection. In addition, a fraction of larger particles (droplets) expelled by coughing or
sneezing will shrink in size by evaporation and hence will become aerosols (droplet nuclei) thus
increasing the total aerosol content. However, 99.9% of the fluid volume is contained in larger
droplets and, as a consequence, this is also where an equivalent proportion of emitted pathogens
would be. This supports the argument that the most likely route of infection would be large
droplets, or subsequent contact with them after they have settled onto surfaces. However, this
does not necessarily mean that infection by inhalation is highly unlikely. This probability is
affected by a number of parameters, including the infectious dose and titre of the pathogen in
secretions. Both of these parameters are unknown for future pandemic influenza strains.
Unlike exposure to industrial chemicals whereby Workplace Exposure Levels (WELs) exist,
there is no specified safe exposure level for biological agents. Whilst the numbers of organisms
required to establish different infections varies, the general requirement is to reduce the
exposure to as low as reasonably practicable. HSE’s current stance is that where there is a
respiratory risk of infection use of FFP3 devices represents best practice, and where these are
not available then FFP2 may be an acceptable, pragmatic compromise. Hence, HSEs guidance
on working with poultry suspected of being infected with highly pathogenic avian influenza
prescribes the use of FFP3 respirators (HSE 2005b) due to the absence of an effective vaccine
and the severity of the consequences relating to exposure. This also acknowledges the widely
held belief that H5N1 avian influenza is transmitted to humans in airborne particles, although it
is thought that a large dose is required. Whilst human-to-human transmission of the H5N1
avian influenza virus has not been observed, for the virus to adapt to growth in humans and
cause a pandemic, there would be a prerequisite for the infectious dose to become reduced. The
precise impact of the pandemic virus is difficult to predict, but levels of morbidity and mortality
are expected to be considerably higher than for seasonal influenza outbreaks (NHS 2005; DH
2007a). Furthermore, the relative properties of transmission of this virus may also differ from
those that cause outbreaks of seasonal influenza.
The issue of protecting healthcare workers against infectious respiratory viruses was highlighted
following outbreaks of Severe Acute Respiratory Syndrome (SARS) in the Far East and Canada
in 2002-2003 (reviewed by Gamage et al., 2005). Significant numbers of healthcare workers
were infected with the virus that causes SARS (Hogg & Huston, 2006; Loeb et al., 2004; Ofner-
Agostini et al., 2006; Seto et al., 2003), which is also thought to be spread primarily via large
4
droplets and direct contact, similarly to Influenza virus (CDC 2005). Retrospective studies on
the clinical attack rates of SARS during the management of outbreaks in the hospital setting
suggested that surgical masks afforded some protection, but this was not enough to significantly
reduce the risk of infection. The studies also suggested that N95 respirators may have provided
added protection (Loeb et al., 2004; Seto et al., 2003) and they, or better RPE, may be required
to adequately protect healthcare workers from SARS (Gamage et al., 2005). Consequently,
N95-rated respirators are the minimum recommendation for healthcare workers who are in
proximity to SARS patients due to the residual risk of aerosol transmission (CDC 2005). It is
reasonable to assume that the same characteristics may apply to outbreaks of influenza, and the
pandemic planning authorities in the USA make similar recommendations for work with
pandemic influenza patients (CDC 2006). The USA APF for N95 is similar to the APF of 10
for UK FFP2. A recent published study on the performance of N95 filtering facepieces found
that the workplace protection factor (WPF) for microorganisms with a mean aerodynamic
diameter <5um were less than the APF of 10, concluding that even N95 respirators are
inappropriate for protection against infectious bioaerosols (Lee et al., 2005).
Epidemics of seasonal influenza occur frequently and explosive outbreaks are not uncommon in
hospitals (reviewed in Salgado et al., 2002). Clearly, healthcare workers are at increased risk of
contracting influenza during such an outbreak, and this can be extrapolated to the situation
expected during a pandemic. As no immunity to the pandemic virus will exist, increased attack
rates, mortality and hospital admissions are expected. It is predicted that 25% of the UK
population will be affected during a pandemic (as compared to 5-10% during seasonal influenza
epidemics), resulting in 50,000 excess deaths and an increase in hospital admissions of 25%
(NHS 2005). It is anticipated that excessive strain will be placed on front-line healthcare and
community services and further disruption would also be expected if healthcare workers become
ill. Some evidence indicates that unvaccinated healthcare workers do contract influenza during
seasonal outbreaks with attack rates exceeding those of the rest of the population (Elder et al.,
1996; Horcajada et al., 2003; Salgado et al., 2002). Furthermore, vaccination of healthcare
workers against seasonal influenza probably reduces the number of recorded absence days
(Wilde et al., 1999).
Similar infection control measures to those used in nosocomial outbreaks of influenza would be
introduced during a pandemic. This will probably include isolation of infected patients in a
dedicated private room or cohort ward, the use of droplet precautions (including PPE and
increased levels of hygiene) that includes the wearing of surgical masks. Historically, surgical
masks were worn by healthcare workers for the purpose of reducing the likelihood of
nosocomial infection in patients and have been shown to be effective for the retention of
expelled droplets originating from the wearer (Inouye et al., 2006, Weber et al., 1993).
However, their value in specifically protecting a worker from an airborne infection hazard is
less well defined.
The Department of Health Pandemic Influenza Scientific Advisory Group recently reviewed the
evidence base for the use of facemasks during an influenza pandemic (DH 2007b). This
includes the use of surgical masks in various settings, including their use by healthcare workers.
Much work has been done to characterise RPE protection against non-biological, or inert,
particles and some of the early work was used to establish the tests that form the basis for the
European Standard for RPE. Several studies have demonstrated that the efficiency of surgical
masks against inert airborne particles is greatly reduced compared to FFP respirators (Derrick &
Gomersall, 2005; Derrick et al., 2006; Lawrence et al., 2006; Tuomi, 1985). This difference in
performance appears to be due to the fact that surgical masks are not sealed to the wearer’s face
as artificially sealing them can increase their efficiency (Derrick et al., 2006; Weber et al.,
1993).
5
Little work has been done to evaluate the level of protection afforded by surgical masks against
bioaerosols (reviewed by Rengasamy et al., 2004). The physical characteristics of inert aerosols
and bioaerosols are generally considered to be comparable. However, extrapolation of these
observations to infectious bioaerosols is complicated by many factors, including infectious dose,
the amount of the organism present and its viability in particles of different sizes. Furthermore,
most studies using bioaerosols where the performance of RPE and/or surgical masks has been
compared have concentrated on the filtration efficiency of the material against aerosolised
bacteria, rather than the overall protection afforded to the wearer (for example, Chen et al.,
1994; McCullough et al., 1997; Qian et al., 1998). The filtration efficiency of surgical masks
was also tested using a surrogate virus, MS2 bacteriophage (Balazy et al., 2006). Filtration
efficiencies of up to 97% for bacteria and up to 85% for MS2 bacteriophage are observed but
these results might be misleading, as they do no account for facial seal leakage. A good facial
seal appears to be key to the overall performance of a mask (Qian et al., 1998, Yassi et al.,
2005), a feature not inherent to surgical masks.
Dreller et al measured the performance of a range of surgical masks against both an inert
sodium chloride aerosol as specified in BS EN 149:2001, and against Staphylococcus aureus.
The authors concluded that the filter efficiencies of the materials used in the construction of the
surgical masks and the face seal performance against a sub-micron particle challenge were
significantly lower than the requirements for FFP1. The filtration efficiencies were also
significantly lower than that claimed by the surgical masks’ manufacturers. The authors also
concluded that a minimum of 40% reduction in exposure to S. aureus was measured on surgical
masks (Dreller et al., 2006). This might equate to a reduction factor of 2.2, although the report
does not state whether the masks were sealed to the test head or not.
Some work was done by HSL a number of years ago to test the effectiveness of some RPE and
surgical masks against a bacterial aerosol challenge (Crook B, Brown RC, Wake D, Redmayne
AC. Final Report; Efficiency of respiratory protective equipment against microbiological
aerosols; IR/L/M/96/05). This demonstrated the poor performance of some surgical masks
against bioaerosols. Despite this, there is a lack of scientific evidence in regard to the protective
effect of surgical masks against infectious aerosols (with reference to worker safety) to support
HSE’s pandemic planning activities. The following study aims to provide further scientific
evidence using a combination of challenge studies to measure the efficiency of disposable FFP
respirators and surgical masks to inert airborne particles. This is related to similar evaluations
using a live influenza virus challenge to surgical masks.
6
2 MATERIALS AND METHODS
2.1 INERT PARTICLE TESTS ON TEST SUBJECT
2.1.1 General
The performance of a range of respirators and surgical masks against an inert aerosol was
measured. The inert particle challenge was designed to simulate the particle size range and
particle number found in a cough/sneeze-generated aerosol of saliva.
The original plan was to test the performance of the respirators and surgical masks when fitted
to a ‘Sheffield’ dummy test head attached to a breathing simulator. This dummy head is
described in BS EN 136:1998 and is employed in testing of various types of RPE. However, the
initial tests showed that it was impossible to achieve a fit comparable with that obtained by an
actual wearer on the dummy head. Alternative test heads were investigated, but improved fits
were still not possible. When performing tests in accordance with British RPE Standards, it is
permitted to aid the sealing of the mask on the dummy head with sealing tape, putty etc. For
this study, sealing of the mask to the test head was not practicable, as this would have prevented
the measurement of leakage around the mask. It was decided that, instead of using a test head
and a breathing simulator, a volunteer test subject would be used.
The pulsed spray was synchronised with the inhalation breath of the test subject. The test
subject was asked to remain still during the test. A sample of the air inside the mask was taken
from a position between the mouth and nose using a sample probe as recommended in the HSE
guidance on fit testing (OC 282/28).
2.1.2 Inert aerosol generation
The aerosol was generated by a simple pulsed compressed air atomiser, fed with a solution of
artificial saliva, manufactured according to the recipe in BS 7115:1988 Part 2. This particular
recipe was chosen as it was an innocuous mix and therefore suitable for use with a human
volunteer. The atomiser generated a poly-dispersed aerosol covering a size range <1µm to
>200µm with approximately 50% of the particle number distribution <20µm and 10% >100µm.
This compares well with the particle size distribution of a cough calculated by Nicas et al
(Nicas et al., 2005), where approximately 50% of the particles were <20µm and 25%>100µm.
The duration of the pulse was approximately 0.5 second. The number of particles generated by
this method was >250,000 particles per ml (measured for the range <10µm). However, the
number of particles within the size range of the instrumentation employed in the inert aerosol
tests (0.02µm to approx. 1.2µm) was approximately 100,000.
2.1.3 Performance measurement
2.1.3.1 TSI Portacount
A TSI Portacount Plus ambient particle-counting device was employed to measure the
effectiveness of the respirators and surgical masks against the generated aerosol. This device is
used extensively in the UK and the USA for fit testing of tight-fitting facepieces (filtering
7
facepieces, half and full face masks). The Portacount determines the degree of fit by comparing
the particle concentration on the outside of the facepiece to the particle concentration inside the
facepiece. The Portacount operates within the size range 0.02µm to approx. 1.2µm. Therefore,
only this size fraction of the generated aerosol (‘sneeze’) is measured and the inert particle
challenge results are based upon this.
In order to determine a measure of face fit only (i.e. the measure of leakage around the facial
seal) penetration of particles through the filtering media of the facepiece and within the size
range of the Portacount has to be eliminated. When testing high efficiency filtering facepieces
(FFP3), the particle penetration through the filter media is negligible and the output of a
Portacount fit test is therefore a measure of the face fit. However, when fit testing masks of
lower particle filtering efficiency such as FFP2, FFP1 and surgical masks, the particle
penetration through the filter media is not negligible and in order to eliminate these particles a
device referred to as an ‘N95 Companion’ is used in conjunction with the Portacount. The ‘N95
Companion’ is a particle classifier that removes from the particles sampled by the Portacount,
the particle size range that is assumed to penetrate lower efficiency filter media. The resulting
particle size range that is measured is approximately 0.03µm – 0.06µm. Therefore by employing
both the Portacount and the Portacount together with the N95 Companion, a measure of both the
face fit and the face fit plus filter penetration can determined.
Therefore in summary:
• Tests performed using the Portacount only – measures face fit and filter penetration.
The results from these tests are reported as reduction factors.
• Tests performed using the Portacount and the N95 Companion – measures face fit only.
The results from these tests are reported as fit factors.
2.1.3.2 Test procedure
A test subject fitted the respirator or surgical mask to be tested in accordance with the
manufacturer’s instructions. Adjustments were made in order to obtain the best fit possible.
For the respirators the target was to achieve, during normal breathing, at least a fit factor of 100
for FFP3 with the Portacount, and 100 for FFP2 and FFP1 with the Portacount with the N95
Companion fitted. A fit factor of 100 was chosen as this is deemed to represent an adequate fit
and is the pass criteria for FFPs published in HSE guidance OC282/28 (HSE 2004) Where the
respirator was judged not to fit, the best-fit possible was obtained and the fact noted. Another
respirator was not chosen, as the purpose of the tests was to test a sample of the products
available from the NHS Logistics.
The volunteer was positioned at a distance of 1m away from the aerosol generator, such that the
aerosol sprayed over the face of the volunteer.
The following tests and measurements were undertaken:
• With the test subject sat still, the performance of the respirator or surgical mask was
measured using the naturally occurring ambient airborne particles only as the challenge
• With the test subject sat still, the performance of the respirator or surgical mask was
measured using an aerosol spray as the challenge. The spray was synchronised with the
test subject’s inhalation breath. The test subject was exposed to at least five sprays
8
For both the above conditions the test was performed with the Portacount only and then with the
Portacount with the N95 Companion fitted to give a measure of face fit and then a measure of
face fit plus filter penetration; the latter is referred to as the Reduction Factor. The test sequence
was randomised to reduce system bias.
Following a complete test the respirator or surgical mask was removed. The test was repeated
twice to give three measurements. As surgical masks are not designed for reuse, a new mask
was used for each test.
A schematic diagram of the test arrangement is shown in Figure 2.1.
Figure 2.1. Schematic diagram of the inert aerosol test arrangement
2.2 RESPIRATORS AND SURGICAL MASKS EMPLOYED IN THE STUDY
A selection of respirators of class FFP1, FFP2 and FFP3 and a selection of surgical masks from
the NHS list of products available via NHS Logistics were tested. The products chosen were
those that had the highest sales for the year 2005-2006. In addition to the highest selling
products others were selected in order to cover a representative sample of the range of masks
available. A copy of the list of products available via NHS Logistics for the year 2005-2006
and the revised list updated 2006 are appended to this report (Sections 7.1 and 7.2). The range
of filtering facepieces available is somewhat limited, and therefore in addition to those chosen
from the NHS lists, other FFP3, FFP2 and FFP1 that are popular ‘industrial’ type respirators
were selected for comparison purposes and to increase the range of products tested.
9
The surgical masks can be crudely subdivided into two main types – those with tie-fastenings,
and those with elastic straps (either single or double). A diagram of the different mask designs
included in the test is shown in Figure 2.2.
Figure 2.2. A diagrammatic representation of those surgical mask designs

included in the tests; I: Typical tie mask (Tie A, B, D and E). II: ‘Duckbill’ strap
mask (Strap A and B). III: Tie mask with integral splash visor (Tie C). IV:
Moulded strap mask (Strap C).
2.3 INFLUENZA BIOAEROSOL TESTS
2.3.1 Virus Culture and Processing
A variant of Influenza virus A/PR/8/34 that was adapted for growth on Madin-Darby Canine
Kidney (MDCK) cells was obtained from the American Type Culture Collection (LGC
Promochem Ltd UK). MDCK cells were obtained from the European Collection of Cell
Cultures (ECACC) and cultured as directed by the supplier. High titre stocks of virus were
grown on cultured MDCK cells essentially as described elsewhere (Gaush & Smith, 1968).
Briefly, cells were inoculated with a 1:500 dilution of virus in serum-free Virus Infection
Medium. After allowing virus to adsorb for 1 hour at 35°C in a humidified 5% CO2 incubator,
the inoculum was removed and cells were maintained in Virus Infection Medium at 35°C.
Virus titre in the supernatant was monitored by Haemagglutination (HA) assay as described
(Cann(Ed), 1999) using chicken red blood cells (TCS Biosciences). When the HA titre was at a
maximum (usually day 3 or 4 post infection) cellular debris was removed from the crude virus
10
preparation by centrifugation at 1000 xg. This clarified viral preparation was subsequently
stored at -80ºC.
Prior to use in surgical mask challenge studies, influenza virus was concentrated from the crude
preparation by ultracentrifugation at 100,000 xg for 2.5 hours at 4ºC. The supernatant was
aspirated and the viral pellet resuspended in Phosphate Buffered Saline (PBS) + 0.2% Fraction
V Bovine Serum Albumin (BSA) overnight at 4ºC. Virus collected from approximately 36ml of
crude preparation was resuspended in 6ml PBS + 0.2% BSA. The presence of virus and a crude
titre was confirmed using HA assay.
2.3.2 Inert aerosol sampling
Prior to the live influenza bioaerosol exposure tests and for each surgical mask tested, a measure
of the reduction factor against the inert aerosol generated using PBS + 0.2% BSA solution (in
which influenza virus was to be suspended) was obtained. The purpose of this test was to
provide a link between the inert aerosol tests on the human volunteer and the live bioaerosol
tests on the dummy head.
The dummy head was positioned within a 1200mm Class II Microbiological Safety Cabinet
immediately opposite a pulsed compressed air atomiser. A diagram of the testing rig is shown
in Figure 2.3. The sampling distance from the atomiser to the front of the dummy head was
approximately 70cm (this is a slightly shorted distance to that employed in earlier tests due to
the size restrictions of the cabinet). A surgical mask was fitted to the dummy head, taking care
to achieve the best representative fit possible. The cabinet airflow was switched off and the
breathing simulator was activated, allowing the head to inhale/exhale at 40 litres per minute
(stroke volume of 2.0 litres x 20 cycles per minute). This breathing rate represents a medium-
low work rate. A 0.5s pulsed spray of PBS + 0.2% BSA was synchronised with the inhalation
breath and air was sampled from both inside and outside the mask using a TSI Portacount
ambient particle counting device to measure the effectiveness of the mask against the inert
aerosol generated. Only the Portacount and not the Portacount plus the N95 Companion was
employed because the measurement of interest in these particular tests was the overall
protection (reduction factor) afforded by the masks and not just the quality of the face fit.
Because the solution in which the influenza virus was suspended was a different solution to that
used in the inert aerosol tests on the human volunteer, the size distribution was re-measured.
The particle size distribution ranged from <1µm to >200µm with approximately 50% of the
particle number distribution <60µm and 15%>100µm. This size distribution of this spray covers
the same range but contains a slightly higher proportion of larger particles than that used in the
tests on the human volunteer.
2.3.3 Live influenza bioaerosol sampling
The cabinet airflow was switched on to enable safe manipulation of the influenza test
suspension. The PBS + 0.2% BSA drained from the atomiser, which was then charged with
concentrated influenza test suspension. Air was sampled directly into virus infection medium in
separate collecting devices concurrently from both immediately in front and immediately behind
each surgical mask. These were connected to a vacuum pump and calibrated to a flow rate of 1
litre per minute in situ using a rotameter.
With the airflow switched off, three separate external and internal samples from the air were
taken for each mask. External and internal negative control air samples were then taken in the
11
absence of a spray of influenza test suspension after bioaerosol sampling (this would also reveal
any cross contamination problems arising during manipulations between samples).
2.3.4 Sample processing
The titre of influenza virus present in the air samples was determined by plaque dilution assay
(Cann(Ed), 1999; Gaush & Smith, 1968). 1/10 and 1/100 dilutions of each sample were
prepared in virus infection medium. Monolayers of MDCK cells grown in 6-well dishes were
inoculated with 250µl of neat sample, or diluted samples. Internal and external sample pairs
were assayed using separate wells of the same dishes. The titre of the influenza test suspension
was also determined using the same assay, except with serial 10-fold dilutions of each sample
were prepared in virus infection medium down to 1x10-15.
Following adsorption for one hour at 35ºC, a molten agarose overlay was prepared which was a
1:1 mix of 3% low gelling temperature agarose and 2x Minimal Essential Medium (see section
7.3 for full constituents). The inoculum was removed and the cells washed in PBS before the
addition of 2ml of molten agarose overlay to the cells. The prepared assay dishes were
incubated at 35ºC for 72 hours.
The overlay prevents the diffusion of viral particles through the medium, allowing the formation
of discrete viral foci of infection, or plaques. After this incubation, viral plaques were counted
in the wells where clearly separate foci could be distinguished and enumerated. The influenza
virus titre could then be expressed as plaque-forming units (PFU) per ml of inoculum.
Figure 2.3. Diagrammatic representation of the influenza bioaerosol testing rig
12
3 RESULTS
3.1 INERT PARTICLE EXPOSURE TESTS ON TEST SUBJECT
3.1.1 Calculation of the reduction factor and fit factors
The performance of the respirators and surgical masks was determined by calculating the ratio
of the particle concentration inside and outside the facepiece as shown below:
Particle concentration outside the facepiece (Co)
Reduction factor = ——————————————————
Particle concentration inside the facepiece (Ci )
The fit factors, i.e. the measure of the leakage around the facial seal, is similarly calculated as
shown above, for tests where the N95 Companion was employed.
For the ambient challenge tests, the mean particle count inside and outside of the facepiece over
a sample period of 1 minute was calculated and used to determine the reduction factor. For the
simulated sneeze, the mean of the peak particle count both inside and outside of the facepiece
for each spray was calculated and used to determine the reduction factor. Figure 3.1 shows an
example of the particle count trace during the test.
Figure 3.1. Example of challenge and in-mask particle concentration during

testing of a surgical mask
13
Each of the respirators and surgical masks were tested three times using three separate fittings.
The means of the three tests were calculated and the results for the Reduction Factors are
presented in the Table 2, Figure 3.2 and Figure 3.3. The results for the Fit Factors are presented
in Table 2 and Figures 3.3 and Figure 3.4. Furthermore the results of the mean of the filtering
facepiece classes and of the surgical masks, separated by design i.e. straps or tie, were
calculated and are presented in Table 3 and Figure 3.6.
Table 2. Reduction factors and fit factor results for the range of filtering
facepieces and surgical masks tested
Reduction Factor Fit Factor
Mask Type Ambient Particles Simulated Sneeze Ambient Particles Simulated Sneeze
FFP3 A 12166 1493 7068 3809
FFP3 B 280 130 800 253
FFP3 C 1559 1480 5520 2941
FFP3 D 91 90 2496 522
FFP3 E 151 70 220 43
FFP2 A 57 26 98 23
FFP2 B 88 85 1925 444
FFP2 C 382 183 1070 84
FFP2 D 282 122 493 187
FFP1 A 94 94 2195 214
FFP1 B 17 27 1513 766
Tie A 3 2 3 2
Tie B 7 4 7 3
Tie C 7 2 12 4
Tie D 2 1 2 1
Tie E 6 3 7 3
Strap A 17 17 1754 358
Strap B 7 4 273 172
Strap C 2 1 3 1
14
REDUCTION FACTOR FOR INERT PARTICLE CHALLENGE
Ambient Particles Simulated Sneeze
FFP3
100000
FFP2
Reduction Factor
10000
FFP1
1000
Surgical
100
10
1
FFP3 FFP3 FFP3 FFP3 FFP3 FFP2 FFP2 FFP2 FFP2 FFP1 FFP1 Tie A Tie B Tie C Tie D Tie E Strap Strap Strap
A B C D E A B C D A B A B C
Type of mask
Figure 3.2. Inert particle reduction factors for all filtering facepieces and surgical
masks tested. Error bars show the spread of results measured
REDUCTION FACTOR FOR INERT PARTICLE CHALLENGE
100
Reduction Factor
10
1
Tie A Tie B Tie C Tie D Tie E Strap A Strap B Strap C
Surgical mask
Figure 3.3. Inert particle reduction factors for all the surgical masks tested.
Error bars show the spread of results measured
15
FIT FACTOR FOR INERT PARTICLE CHALLENGE
FFP3
100000
FFP2 FFP1
Surgical
10000
Fit Factor
1000
100
10
1
FFP3 FFP3 FFP3 FFP3 FFP3 FFP2 FFP2 FFP2 FFP2 FFP1 FFP1 Tie A Tie B Tie C Tie D Tie E Strap Strap Strap
A B C D E A B C D A B A B C
Type of mask
Figure 3.4. Fit factors for all filtering facepieces and surgical masks tested.
Error bars show the spread of results measured
FIT FACTOR FOR INERT PARTICLE CHALLENGE
10000
1000
Fit Factor
100
10
1
Tie A Tie B Tie C Tie D Tie E Strap A Strap B Strap C
Surgical mask
Figure 3.5. Fit factors for all the surgical masks tested. Error bars show the
spread of results measured
16
Table 3. Harmonic mean values of the reduction factors and fit factor results for
the grouped range of filtering facepieces and surgical masks tested
Reduction Factor Fit Factor
Mask Type Ambient Particles Simulated Sneeze Ambient Particles Simulated Sneeze
FFP3 228 145 766 167
FFP2 95 54 258 52
FFP1 29 42 1791 335
Surgical mask - tie 4 2 4 2
Surgical mask - strap 5 2 9 2
Surgical mask - all 4 2 5 2
Note: When calculating the ‘mean’ of a series of fit factors or reduction factors, the correct method is to calculate the
harmonic mean, which is the reciprocal of the arithmetic mean of the reciprocals of the data series.
MEAN REDUCTION FACTOR AND FIT FACTORS AGAINST INERT PARTICLE

CHALLENGE
Reduction Factor Simulated Sneeze Fit Factor Simulated Sneeze
1000
Reduction Factor / Fit Factor
100
10
1
FFP3 FFP2 FFP1 Surgical mask
Type of mask
Figure 3.6 Mean values of the reduction factors and fit factor results for the
grouped range of filtering facepieces and surgical masks tested against the inert
simulated sneeze
17
3.1.1.1 Reduction factors during exposure to the simulated sneeze
As expected, the performance provided by the filtering facepieces increased progressively from
FFP1 to FFP3. The reduction factors for the filtering facepieces when exposed to the simulated
sneeze ranged from 25 to >4000.
The performance provided by all the seven surgical masks tested against the simulated sneeze
were significantly lower than the minimum requirement for a class FFP1 filtering facepiece.
The reduction factors achieved with each mask when exposed to the simulated sneeze ranged
from 1.0 to 16.5. A reduction factor of 1 represents zero protection. All surgical masks with
ties returned a reduction factor <10. One out the three surgical masks fitted with elasticated
straps achieved a reduction factor >10.
3.1.1.2 Fit factors during exposure to the simulated sneeze
The fit factors for the filtering facepieces ranged from 25 to >3000; a fit factor >100 is deemed
a satisfactory fit (OC282/28). During the tests, when only the ambient airborne particles were
challenging the filtering facepieces, only one facepiece returned a fit factor <100. This
moulded-cup shaped facepiece was deemed by the test subject to ‘not fit’, and therefore would
have been rejected when selecting suitable RPE. This type was not on the NHS Logistics list of
products. When subjected to the simulated sneeze (i.e. a particle challenge concentration 10
fold that of the ambient particle challenge concentration) two further filtering facepieces
returned a fit factor <100.
The fit factors measured on the five surgical masks with ties were significantly lower than 100.
The maximum fit factor measured against the ambient particle challenge was 12 and the lowest
was 2.0. The fit factors against the simulated sneeze were lower at a maximum of 4 and a
minimum of 1.0. The surgical masks with straps fared better with two out of the three tested
returning fit factors >100 under ambient and simulated sneeze conditions. The fit of the
surgical mask ‘Strap-C’ fitted with only one elasticated strap was extremely poor, returning a fit
factor of 1.0.
3.2 INFLUENZA BIOAEROSOL TESTS
Quantitative demonstrations of the relative levels of protection afforded by surgical masks

against live aerosolised influenza virus in a simulated sneeze were conducted. The range of
seven surgical masks used in the inert particle tests was evaluated.
All masks were tested in accordance with the procedure set out in Section 2.3.3. The testing rig
is shown in Figure 2.3. The performance of the surgical masks with respect to the influenza
bioaerosol was determined by calculating the ratio of live virus sampled from the air outside
versus that sampled from inside the mask, using a plaque assay. This ratio was termed the
“influenza plaque reduction factor” i.e:
Influenza virus titre of external air sample
Influenza plaque reduction factor = —————————————————
Influenza virus titre of internal air sample
18
Therefore, the influenza plaque reduction factor is broadly equivalent to the reduction factor
measured for the inert particle tests calculated using the Portacount equipment (except that for
the inert particle tests, only the particle sizes within the Portacount’s range were measured).
The inert particle reduction factor is an estimate of the level of fit and filter penetration achieved
on the dummy head in the experimental setting.
The results for all surgical masks tested are shown in Figure 3.7. The inert particle reduction
factors achieved with each mask varied greatly, ranging from 1.3 to 20. Likewise, the influenza
plaque reduction factor varied between masks, ranging between 1.1 and 55. The performance of
the surgical masks in the inert particle challenges was related to the performance against
influenza virus bioaerosols – i.e. a higher performance in inert particle tests was associated with
better performance in the bioaerosol assays. Apart from two notable exceptions (masks “Tie
C” and “Strap A”, which perform better in these tests), all the masks demonstrated that they
would reduce the exposure to infectious influenza virus present in a direct challenge by around
10-fold or less (i.e. a 1 log reduction or below). The calculated harmonic mean influenza plaque
reduction factor for all bioaerosol challenge tests was 6.
INFLUENZA VIRUS CHALLENGE TO SURGICAL MASKS
Influenza Plaque Reduction Factor Inert Particle Reduction Factor
100
Reduction Factor
10
1
Tie A Tie B Tie C Tie C II Tie D Tie E Strap A Strap A II Strap B Strap B II Strap C
Surgical Mask
Figure 3.7. Mean (harmonic) influenza plaque reduction factors and associated
inert particle reduction factors for all surgical masks tested. Error bars show
spread of the calculated data
19
4 DISCUSSION
The aim of this study was to evaluate the ability of surgical masks to provide respiratory
protection against an influenza virus bioaerosol and compare this to the performance of FFP
respirators. A combination of challenge studies was used to assess the performance of surgical
masks and FFP respirators to non-biological airborne particles and surgical masks to an aerosol
of influenza virus.
4.1 FITTED FACEPIECE RESPIRATORS
As can be seen from the fit factor results shown in Figure 3.4, measures of fit achieved for FFP2
and FFP1 were sometimes better than that measured for FFP3. There are two possible reasons
for this:
• Firstly, the particular design (size & shape) of the FFP1/2 may be more suitable to the
test subject, for example FFP2-D which has two separate elasticated straps, together
with a longer nose seal did fit the test subject much better than the FFP3-E facepiece,
which has a short nose strip and a single looped elasticated strap.
• Secondly, the breathing resistance increases as more filtering layers are added to the
facepiece in order to achieve higher efficiencies. The higher the breathing resistance
the higher the possible face seal leakage. If a face seal leakage is present, then the
proportion of leakage via this route increases proportional to the breathing resistance.
For example Fit Factors for FFP1-A, FFP2-C and FFP3-B, decreased in ascending order
of class (i.e. FFP1-A>FFP2-C>FFP3-B) as increasing layers of filter media are used.
In most cases, FFP3 facepieces have additional sealing aids over that of lower classes of
filtering facepieces, such as a rubber/foam face seal around the facepiece. FFP3s that have such
sealing aids (e.g. FFP3-A and FFP3-C) tend to perform better than FFP2s. This was also
evident in this study. Neither of these filtering facepieces is on the list of products provided via
NHS Logistics.
4.2 SURGICAL MASKS
The model scenario was a viral bioaerosol generated by a cough or a sneeze from a potentially
infectious patient. The respiratory protection provided by the surgical masks against an inert
challenge was significantly lower than that provided by the FFPs. The reduction factors and fit
factors measured for the surgical masks with ties failed to achieve a reduction factor of 10.
Whilst the surgical masks with two elasticated straps (Strap A and Strap B) fitted better than the
surgical masks with ties and achieved an order of fit that would be acceptable as a FFP (i.e.
>100; see Figure 3.4), the filtration efficiency was still very low and considerably lower than the
results obtained for the FFP1 (see Figure 3.1).
Whilst the physical characteristics of inert aerosols and bioaerosols are generally considered to
be comparable, direct extrapolation of these observations to infectious bioaerosols would not
necessarily be accurate. For example, the ability of an organism to remain viable in an aerosol
is an important consideration. If the organism cannot survive in an aerosol long enough, the
effectiveness of a surgical mask against influenza virus might be higher than that indicated by
20
inert particle challenges. Therefore we have undertaken both inert particle tests alongside
similar analyses using a live, representative microbiological challenge.
Influenza virus was the chosen agent for this study. This is particularly relevant as the UK
makes preparations for a potential pandemic of influenza. An A-type influenza virus (strain
A/PR/8/34, subtype H1N1) was used, which is expected to have similar biophysical properties
to those that have caused previous pandemic outbreaks, as well as the current H5N1 candidate
strain.
A dummy head, attached to a breathing simulator and wearing a surgical mask, was subjected to
a direct challenge with an aerosol containing live, infectious influenza virus. The distance from
the point at which the aerosol was generated to the outside of the surgical mask was 70cm (i.e.
within one metre). Air was sampled concurrently, but separately, from immediately in front of
the mask and immediately behind the mask over a short period (1 minute). The quantitative
assay performed on the sampled air only measured levels of viable virus. Thus an evaluation of
the protective effect of surgical masks against a direct challenge with an influenza bioaerosol
was possible. Furthermore, the data could be compared to that obtained using inert particle
challenges, which model the physical characteristics of the scenario more closely.
The original intention was to test both surgical masks and FFP respirators but initial tests
showed that it was difficult to achieve a good fit on the ‘Sheffield’ dummy head (see Section
2.1). Artificially sealing respirators or surgical masks to the test head was not considered
representative of the model scenario. It was decided to limit this phase of the study to the
analysis of surgical masks only, as the level of fit achieved on the dummy head was comparable
to that achieved on the human test subject during inert particle challenge tests. Inert particle and
bioaerosol tests were conducted in parallel on the surgical masks. Since the performance of the
surgical masks in the inert particle challenges was related to that against bioaerosols, some
extrapolation of the data is now possible.
These tests were performed in a closed microbiological safety cabinet. In reality, infectious
bioaerosols would be generated in more open space, therefore diluting the infectious organism
over time and reducing the likelihood of infection. Therefore, it was not possible to directly
mimic this dilution effect. There are also environmental parameters that may affect the viability
of the virus and the characteristics of the bioaerosol (e.g. temperature, humidity, ventilation etc)
that were not accounted for in this study. However, air was only sampled for a short period
(one minute) so the effects of dilution and environmental factors would not be expected to
substantially impact upon the results obtained using these tests. Furthermore, the influenza
bioaerosol interacts with the external sampler more directly. Therefore, if there is any bias in the
sampling system, it would artificially increase the proportion of influenza virus recovered by the
external sampler, resulting in a higher influenza plaque reduction factor and perceived
protective effect.
Live, infectious virus was extracted in enumerable quantities from the air from behind all the
surgical masks tested. This suggests that influenza virus can survive in aerosol particles and
bypass/penetrate a surgical mask and that a residual infectious aerosol hazard may exist.
Whether or not the surgical masks tested will offer adequate protection against infection will be
dependent on the infectious dose of the virus, and its titre in secretions. The infectious dose for
humans with respect to a potential pandemic virus is unknown and cannot be determined until
the virus itself emerges. The 50% human infectious dose for another strain of influenza has
been estimated at 0.6 to 3.0 50% tissue culture infectious doses (TCID50; Alford et al., 1966). If
it is assumed that 1 TCID50/ml equates to approximately 0.5 PFU/ml, then the PFU needed for
human infection will be of similar order. The amount of infectious virus present in nasal
secretions has been estimated to reach levels of 107 to 108 TCID50/ml (Murphy et al., 1973).
21
The concentration of influenza virus in the test samples was between 1014 and 1015 PFU/ml.
Therefore, the levels of virus present in the bioaerosol challenge were probably considerably
higher than those present in a natural cough or sneeze emanating from a pandemic influenza
patient. These levels were required in order to recover sufficient virus to make meaningful
quantitative assessments of surgical mask performance. At these levels, there may have been a
protective effect on the stability of the virus. However, since these effects would be reflected in
the amount of virus recovered in both the internal and external samplers, this would not be
expected to affect the calculated reduction factors.
The performance of surgical mask Tie C was consistently better than other surgical masks in the
bioaerosol challenge and the parallel inert particle tests undertaken using the dummy head.
However this mask did not perform at this level when worn by the human test subject. This
mask has an integral visor, which may have protected the major leakage areas (e.g. around the
nose) and blocked the bioaerosol challenge sufficiently from reaching this area so as to improve
overall protection. Similar masks, or the use of a splash visor in conjunction with a
conventional surgical mask, could be employed to give added protection to the respiratory
system in the event that FFP respirators are not available. The use of a visor would also protect
the eyes from large droplets/splashes and would discourage manual inoculation of the eyes via
direct contact. Further research is needed to evaluate the ability of a visor to enhance protective
effect of a surgical mask.
Most of the surgical masks tested gave a reduction factor in aerosolised influenza of around 1
log, although some were considerably lower than this (the harmonic mean for all bioaerosol
challenge tests was 6 fold). The data presented here demonstrate that, in order to afford a
consistent protective effect above this level, an FFP respirator is required. This is highlighted
by the results of strap mask B (an N95 class respirator) where a good level of fit can be
achieved and performance is similar to FFP1/FFP2 respirators in the inert particle tests with a
human wearer (see Figure 3.4 and Figure 3.5). However, a good level of fit cannot be achieved
on the dummy head, and the mask performs less well in both inert particle and bioaerosol tests
(see Figure 3.7).
The results of the inert particle challenge tests, which are also supported somewhat by the
bioaerosol challenge tests, indicated that surgical masks with double elasticated straps
performed better than the surgical masks with ties. The closer the design of the surgical mask to
that of a FFP respirator (e.g. surgical mask Strap A) the better the fit and therefore the better the
protection afforded. This highlights the fact that correct fitting of masks with a facial seal is
required to achieve the required protection.
Other studies have measured the penetration of various inert and bioaerosols and have shown
that the efficiency of the filter media employed in surgical masks is very variable ranging from
10 to >90% (Chen et al., 1994; Dreller et al., 2006; McCullough et al., 1997; Weber et al.,
1993). Two of these studies investigated face fit of surgical masks and, like the findings of this
study, found the fit factors to be low (Dreller et al., 2006; Pippin et al., 1987). Additional
protection in inert particle challenges has also been demonstrated when the sides of surgical
masks were taped to the face to minimise leakage (Derrick et al., 2006).
Even if the mask is manufactured from high efficiency filtering media, a high proportion of
particles challenging the surgical mask will enter the breathing zone via breaches in the face
seal. Furthermore, a high efficiency filtration media and fluid-resistant layers are likely to
increase breathing resistance. This, together with a poor face fit, will increase the degree of
leakage around the face seal. This is of concern as fluid-repellent surgical masks (typically with
poor face fit characteristics) are being recommended for work in proximity to patients infected
22
with pandemic influenza to protect against splashes and large droplets (HPA 2005; DH 2007a;
NHS 2007).
As surgical masks cannot be fitted well to the face, their use may not be adequate for protection
against a residual airborne infection hazard. Based upon the measurements obtained in this
study, on average, a reduction factor of 10 or below can be anticipated. Whether or not that
would negate the need for widespread use of FFP respirators will depend upon the nature of,
and consequences of exposure to, the organism.
It is important that the distinction between the protection factors assigned (APF) to FFPs and the
results obtained in this study is not overlooked. The APFs are derived from the 5th percentile
point of ranked data obtained from workplace protection factor (WPF) studies following an
accepted protocol and within an effective RPE management programme. APFs cannot be
calculated from the tests performed in this study. However, if the data that has been obtained for
surgical masks against the influenza virus challenge were from accepted workplace studies, the
calculated 5th percentile would be 2.4 – this value for surgical masks would be significantly less
than the harmonic mean of 6 quoted earlier in the report. Although FFP3 were not tested against
the bioaerosol, if a similar relationship existed as for that for the surgical masks between the
results for the inert and bioaerosol challenges, then the use of well fitting FFP3 respirators
would potentially give a reduction factor of at least 100.
23
5 CONCLUSIONS
In principle, surgical masks provide adequate protection against large droplets, splashes and
contact transmission. There is a common misperception that they will provide protection against
aerosols. This study did not assess the protective capacity of surgical masks against large
droplets, splashes and contact but rather focussed on their performance against a respiratory
aerosol challenge. The results of this study confirm that surgical masks provide the lowest level
of respiratory protection compared to FFP respirators. Furthermore, the level of protection
afforded by surgical masks against inert aerosols is similar to the level of protection afforded
against a live bioaerosol containing influenza virus. A reduction factor of around 6 can be
anticipated, depending on the type of surgical mask used. Many surgical masks on the NHS
logistics list tested here offer considerably less protection than this.
It is recognised that a pandemic of influenza will result in a sustained period of pressure on

healthcare service providers, particularly at the point of patient care. Healthcare service
delivery arrangements will be adapted to cope with the rise in demand at all levels, including
patient management and the provision of treatment. These arrangements will involve
organisational, environmental and procedural controls and will probably be based upon normal
delivery mechanisms as far as is practicable. Central to these arrangements is the application of
effective infection control procedures to prevent nosocomial spread of the virus within hospitals
caring for those patients with severe symptoms that require treatment.
The national risk assessment should be able to demonstrate that there is no or negligible risk
arising from naturally occurring respiratory aerosols. Even if the probability of infection is
considered to be low, the consequences, both to the exposed healthcare worker, and to the
ability of healthcare service providers to cope with sustained pressure during a pandemic,
should be carefully considered. If there is a residual airborne risk of harm to health, respiratory
protection may be required.
If it is decided that there is an additional need for respiratory protection, organisational and
management controls may need to be reviewed to allow those healthcare workers required to be
in proximity to infectious patients to be supplied with appropriate respirators. The use of FFP
respirators would necessitate correct maintenance, correct storage, fit testing and use by trained
personnel. Furthermore, tailoring supply of dedicated respirators to individuals may also
represent an issue. These aspects, as well as respirator demand and increased costs, present
planning challenges to the healthcare sector. The widespread use of respirators might be
difficult to sustain during a pandemic unless provision is made for their use in advance.
Surgical masks may provide adequate protection against large droplets, splashes and contact
transmission. They may also reduce any residual aerosol risk but it remains unclear whether this
level of protection sufficiently reduces the likelihood of transmission via this route so as to
minimise the risk of infection to as low as reasonably practicable. With this in mind, it is
recommended that HSE draw the results of this research to the attention of DH/HPA so that it
can be considered as part of the wider issue of modes of influenza transmission.
24
6 RECOMMENDATIONS FOR FUTURE WORK
This study has generated valuable data and further evidence as to the effectiveness of surgical
masks against aerosols. The testing system developed is novel insofar as it models a direct
challenge with live airborne influenza virus to a healthcare worker wearing a surgical mask
more realistically than any previously published study. There were some limitations to the
study, for example, filtering facepieces were not subjected to the influenza bioaerosol tests and
the tests were carried out on only one human test subject or a dummy test head. No attempt was
made to include a representative sample of facial shapes and sizes. Some of these limitations
could be addressed in further studies using the bioaerosol testing system developed here, or by
developing the system further.
6.1 TESTING A BROADER RANGE OF SURGICAL MASKS AND

RESPIRATORS
The surgical mask designs employed in the study were representative of the variety of available
surgical masks, but only a relatively small number of surgical masks were the tested. A broader
range of masks could therefore be tested using the testing system developed here. Inert particle
challenge tests could also be tested on a variety of facial shapes and sizes
6.2 MAXIMISING THE PROTECTIVE EFFECT OF SURGICAL MASKS

AGAINST INFLUENZA BIOAEROSOLS
The results of the bioaerosol challenges indicated that a mask with a built-in splash visor offers
better protection than a conventional surgical mask of similar design and construction but
without the integral visor. This should be further explored by performing inert particle and
influenza bioaerosol tests on masks with built-in fluid shield compared with their conventional
surgical mask counterparts. Also, conventional masks in conjunction with a separate splash
visor could be tested.
The level of added protection achieved by taping the sides of surgical masks to the face in order
to reduce leakage could also be investigated.
6.3 ANALYSIS OF SURGICAL MASK PROTECTION AGAINST A MORE

REPRESENTATIVE INFLUENZA BIOAEROSOL
The levels of virus present in the bioaerosol challenge used in this study were probably
considerably higher than those present in a natural cough or a sneeze. The studies outlined here
could be repeated using influenza virus at levels more representative of a real sneeze. This
would probably not permit quantitative data to be obtained regarding the protection factor
afforded by the masks. However, the ability to recover live, infectious virus from behind the
mask following a lower-titre challenge could be tested.
25
6.4 TESTING OF THE SURVIVAL AND DISSEMINATION OF LIVE
INFLUENZA BEYOND ONE METRE
A larger microbiological safety cabinet or bespoke isolator could be used to test the
effectiveness of surgical masks to influenza bioaerosols at a distance greater than 70cm, or
against an indirect bioaerosol challenge (i.e. an ambient bioaerosol challenge test).
Alternatively, the HSL exposure chamber could be employed to conduct bioaerosol tests in a
side-room mock-up. This would allow the analysis of surgical mask protection beyond a
distance of one metre, as well as allowing for the dilution effect of a larger area. Furthermore,
the ability of live influenza to travel distances of greater than one metre in an aerosol could be
analysed.
6.5 BIOAEROSOL CHALLENGE TO FFP RESPIRATORS
FFP respirators were not tested against bioaerosols in the current study because it was difficult
to achieve a good fit on the dummy head (see Section 2.1). The data presented here enables us
to extrapolate the influenza bioaerosol data from surgical masks to infer the level of protection
likely from an FFP respirator. However, these respirators could be tested directly by artificially
sealing them to the dummy head. This would enable definitive data to be obtained on the
effectiveness of properly fitted FFP respirators against an influenza bioaerosol.
26
7 APPENDICES
7.1 DETAILS OF PRODUCTS AVAILABLE VIA NHS LOGISTICS (LIST

RECEIVED AT HSL FEB MAY 2006)
Masks sales analysis by product

2005/2006 projected sales for financial year (based on 9 months sales)
Projected Projected total

quantity num ber of
Manufacturers purchased by individual units
NPC Code product code Brand Supplier pack size purchased
FFP1 Respirators
Mask face respirator EN149 Class FFP1 Case of
BTP030 42902 Barrier Molnlycke 120
20 6
Mask face respirator Unvalved to EN149 class
BTP009 1861 3M 3M FFP1S Pack of 20 500 10,000
FFP2 Respirators
BTP003 1862 3M 3M FFP2S Pack of 20 8,300 166,000
Mask face respirator Valved to EN149 class
BTP010 1872V 3M 3M FFP2S Pack of 10 1,780 17,800
BTP018 HM240001 Moldex Universal FFP2S Pack of 20 4 80
Mask face respirator EN149 Class FFP2 Case of
BTP031 42904 Barrier Molnlycke 20 15 300
BTP021 UN62408 Tecnol Universal FFP2 and niosh standards (N95) for TB, pack of 50
15,030 751,500
FFP3 Respirators
BTP006 1863 3M 3M FFP3S Pack of 20 1,010 20,200
BTP011 1873V 3M 3M FFP3S Pack of 10 1,270 12,700
BTP017 HM250501 Moldex Universal FFP3S Pack of 20 50 1,000
Surgical face mask, 4 ties

Mask face 4 ties with noseband non woven
BWM005 UN48100 Liteone Universal disposable Pleated Pack of 50 165,250 8,262,500
BWM006 42280 Barrier Molnlycke disposable Pleated Pack of 50 39,850 1,992,500
BWM014 UN48207 Tecnol Universal disposable Pleated with fluid membrane, pack of
50 3,770 188,500
BWM016 1818 3M 3M disposable Blue pleated fluid resistant, pack of 50
2,680 134,000
BWM055 4233 Barrier Molnlycke disposable Pouched, pack of 50 280 14,000
BWM056 UN48247 Tecnol Universal disposable Pleated with plastic fluid shield, pack of
25 15,800 395,000
Face masks, miscellaneous

Mask face surgical aseptic, Blue moulded rigid fluid
BWM022 1800 3M 3M resistant elastic headloop, pack of 50 1,200 60,000
BWM205 6700 Robinson Healthcare
Robinson Healthcare
Mask face cestra muslin Sterile Pack of 36 120 4,320
Mask face particle free Duckbill double elasticated
BWM009 UN47700 Tecnol Universal headband Pack of 50 2,000 100,000
Total quantity sold 258915 12,130,520
27
7.2 DETAILS OF PRODUCTS AVAILABLE VIA NHS LOGISTICS (LIST

RECEIVED AT HSL MAY 2006)
Respirators available via NHS Logistics

Qty Sold
Supplier Brand MPC NPC Description 2005/2006
(singles)
3M 3M 1861 BTP009 Mask face respirator class FFP1 unvalved 10,640
Medisavers Bodyguards FM3015B BTP033 Mask face respirator class FFP2 unvalved (new) new
Molnlycke Barrier 42904 BTP031 Mask face respirator class FFP2 unvalved 1,820
Universal Tecnol UN62408 BTP021 Mask face respirator class FFP2 unvalved and NIOSH standards (N95) 671,550
Medisavers Bodyguards FM3016B BTP034 Mask face respirator class FFP2 valved (new) new
3M 3M 1872V BTP010 Mask face respirator class FFP2 valved 19,770
Medisavers Bodyguards FM3017B BTP035 Mask face respirator class FFP3 unvalved (new) new
Universal Tecnol UN62360 BTP036 Mask face respirator class FFP3 valved (new) new*
3M 3M 1873V BTP011 Mask face respirator class FFP3 valved 20,690
* sales expected to be high
Surgical masks available via NHS Logistics
Qty Sold
2005/2006
Supplier Brand MPC NPC Description (singles)
Molnlycke Barrier 42280 BWM006 Mask face non woven disposable 1,972,600
Molnlycke Barrier 4280 BWM031 Mask face non woven disposable pleated (new) new
Universal Lite One UN48100 BWM005 Mask face non woven disposable pleated 8,400,800
3M 3M 1826 BWM025 Mask face non woven disposable pleated (new) new
Cardinal Cardinal FS71050 BWM029 Mask face non woven disposable pleated (new) new
Molnlycke Barrier 4233 BWM055 Mask face non woven disposable pouched 60,050
Universal Tecnol UN47700 BWM009 Mask face particle free duckbill double elasticated headband 106,950
3M 3M 1838 BWM026 Mask face particle free duckbill with 4 ties (new) new
3M 3M 1818 BWM016 Mask face non woven blue pleated fluid resistant 134,500
Molnlycke Barrier 4234 BWM020 Mask face non woven disposable pleated fluid resistant (new) new
Cardinal Cardinal AT74535 BWM028 Mask face non woven disposable pleated fluid resistant (new) new
Universal Tecnol UN48207 BWM014 Mask face non woven disposable pleated with fluid membrane 197,500
3M 3M 1800+NL BWM022 Mask face surgical blue moulded ridged fluid resistant elastic headloop latex free 63,850
3M 3M 1835FS BWM021 Mask face non woven disposable pleated with fluid shield (new) new
Cardinal Cardinal AT74635 BWM027 Mask face non woven disposable pleated with fluid shield (new) new
Universal Tecnol UN48247 BWM056 Mask face non woven disposable pleated with plastic fluid shield 1,121,900
Molnlycke Barrier 4232 BWM019 Mask face non woven disposable pleated with fluid shield (new) new
28
8 REFERENCES
Alford, R. H., Kasel, J. A., Gerone, P. J. & Knight, V. (1966). Human influenza resulting from
aerosol inhalation. Proc Soc Exp Biol Med 122, 800-4.
Balazy, A., Toivola, M., Adhikari, A., Sivasubramani, S. K., Reponen, T. & Grinshpun, S. A.
(2006). Do N95 respirators provide 95% protection level against airborne viruses, and
how adequate are surgical masks? Am J Infect Control 34, 51-7.
Bell, D. M. (2006). Non-pharmaceutical interventions for pandemic influenza, national and

community measures. Emerg Infect Dis 12, 88-94.
Brankston G, Gitterman L, Hirji Z, Lemieux C, Gardam M (2007). Transmission of influenza A

in human beings. Lancet Infect Dis. Apr;7(4):257-65. Review.
Bridges, C. B., Kuehnert, M. J. & Hall, C. B. (2003). Transmission of influenza: implications

for control in health care settings. Clin Infect Dis 37, 1094-101.
BS 7115: Part 2:1988 Determination of volatile nitrosamines in rubber teats for feeding bottle
and babies’ dummies, Part 2. Artificial saliva extraction method. British Standards
Institute, London.
BS EN 136:1998 Respiratory protective devices: Full face masks - Requirements, testing,

marking. British Standards Institute, London.
BS EN 149:2001 Respiratory protective devices: - Filtering half masks to protect against

particles - Requirements, testing, marking. British Standards Institute, London.
BS EN 529:2005 Respiratory protective devices - Recommendations for selection, use, care and
maintenance - Guidance document. British Standards Institute, London
BS EN 14683:2005 Surgical masks: Requirements and test methods. British Standards Institute,
London
Cann(Ed), A. J. (1999). Virus Culture. A Practical Approach. Oxford University Press.
CDC 2005: Interim Domestic Guidance on the Use of Respirators to Prevent Transmission of
SARS. Available at: http://www.cdc.gov/ncidod/sars/respirators.htm
CDC 2006: Interim Guidance on Planning for the Use of Surgical Masks and Respirators in
Health Care Settings during an Influenza Pandemic. Available at:
http://www.pandemicflu.gov/plan/healthcare/maskguidancehc.html#masks
Chen, S. K., Vesley, D., Brosseau, L. M. & Vincent, J. H. (1994). Evaluation of single-use
masks and respirators for protection of health care workers against mycobacterial
aerosols. Am J Infect Control 22, 65-74.
Derrick, J. L. & Gomersall, C. D. (2005). Protecting healthcare staff from severe acute
respiratory syndrome: filtration capacity of multiple surgical masks. J Hosp Infect 59,
365-8.
Derrick, J. L., Li, P. T., Tang, S. P. & Gomersall, C. D. (2006). Protecting staff against airborne
viral particles: in vivo efficiency of laser masks. J Hosp Infect 64, 278-81.
29
DH 2007a: Pandemic influenza: A national framework for responding to an influenza pandemic.
Available at:
http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAnd
Guidance/DH_080734
DH 2007b: The Use Of Face Masks During An Influenza Pandemic: Scientific Evidence Base.
Available at:
http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAnd
Guidance/DH_077276
Dreller, S., Jatzwauk, L., Nassauer, A., Paszkiewicz, P., Tobys, H.-U. & Rüden, H. (2006). Zur
Frage des geeigneten Atemschutzes vor luftübertragenen Infektionserregern.
Gefahrstoffe - Reinhaltung der Luft 66.
Elder, A. G., O'Donnell, B., McCruden, E. A., Symington, I. S. & Carman, W. F. (1996).
Incidence and recall of influenza in a cohort of Glasgow healthcare workers during the
1993-4 epidemic: results of serum testing and questionnaire. Bmj 313, 1241-2.
Gardam, M. and Lemieux, C. (2007). Transmission of influenza A in human beings - Authors'

reply. Lancet Infect Dis 7: 761-3.
Gaush, C. R. & Smith, T. F. (1968). Replication and plaque assay of influenza virus in an
established line of canine kidney cells. Appl Microbiol 16, 588-94.
Hogg, W. & Huston, P. (2006). Controlling droplet-transmitted respiratory infections: best

practices and cost. Can Fam Physician 52, 1229-32.
Horcajada, J. P., Pumarola, T., Martinez, J. A., Tapias, G., Bayas, J. M., de la Prada, M., Garcia,
F., Codina, C., Gatell, J. M. & Jimenez de Anta, M. T. (2003). A nosocomial outbreak
of influenza during a period without influenza epidemic activity. Eur Respir J 21, 303-
7.
HPA 2005: Infection Control in Hospitals and Primary Care Settings: Pandemic Influenza
Infection Control Guidance. Available at:
http://www.dh.gov.uk/assetRoot/04/12/17/54/04121754.pdf)
HSE 2004: Fit testing of respiratory protective equipment facepieces. HSE 282/28.
http://www.hse.gov.uk/pubns/fittesting.pdf.
HSE 2005a: HSG53 Respiratory protection at work- A practical guide. HSE Books ISBN 0-
7176-2904-X.
HSE 2005b: Avoiding The Risk Of Infection When Working With Poultry That Is Suspected
Of Having Highly Pathogenic Avian Influenza (HPAI). Available at:
http://www.hse.gov.uk/biosafety/diseases/aisuspected.pdf
Inouye, S., Matsudaira, Y. & Sugihara, Y. (2006). Masks for influenza patients: measurement of
airflow from the mouth. Jpn J Infect Dis 59, 179-81.
Lawrence, R. B., Duling, M. G., Calvert, C. A. & Coffey, C. C. (2006). Comparison of

performance of three different types of respiratory protection devices. J Occup Environ
Hyg 3, 465-74.
30
Lee, S. A., Adhikari, A., Grinshpun, S. A., McKay, R., Shukla, R., Zeigler, H. L. & Reponen, T.
(2005). Respiratory protection provided by N95 filtering facepiece respirators against
airborne dust and microorganisms in agricultural farms. J Occup Environ Hyg 2, 577-
85.
Lee R.V. (2007). Transmission of influenza A in human beings. Lancet Infect Dis 7:760-1.
(Author reply 761-3).
Lenhart, S. W., Seitz, T., Trout, D. & Bollinger, N. (2004). Issues affecting respirator selection
for workers exposured to infection aerosols: Emphasis on healthcare settings. App.
Biosafety 9, 20-36.
Loeb, M., McGeer, A., Henry, B., Ofner, M., Rose, D., Hlywka, T., Levie, J., McQueen, J.,
Smith, S., Moss, L., Smith, A., Green, K. & Walter, S. D. (2004). SARS among critical
care nurses, Toronto. Emerg Infect Dis 10, 251-5.
Loosli, C., Lemon, H., Robertson, O. & Appel., E. (1943). Experimental airborne influenza
infection. I. Influence of humidity on survival of virus in air. Proc Soc Exp Biol Med
53, 205-6.
McCullough, N. V., Brosseau, L. M. & Vesley, D. (1997). Collection of three bacterial aerosols
by respirator and surgical mask filters under varying conditions of flow and relative
humidity. Ann Occup Hyg 41, 677-90.
Murphy, B. R., Baron, S., Chalhub, E. G., Uhlendorf, C. P. & Chanock, R. M. (1973).
Temperature-sensitive mutants of influenza virus. IV. Induction of interferon in the
nasopharynx by wild-type and a temperature-sensitive recombinant virus. J Infect Dis
128, 488-93.
NHS 2005: Influenza Pandemic Contingency Planning: Operational Guidance For Health
Service Planners In England. Available at:
http://www.dh.gov.uk/assetRoot/04/11/10/82/04111082.pdf
NHS 2007: Pandemic Flu - A summary of guidance for infection control in healthcare settings.
Available at: http://www.dh.gov.uk/en/PandemicFlu/DH_078752
Nicas M, Nazaroff WW and Hubbard A. (2005). Toward understanding the risk of secondary
airborne infection: emission of respirable pathogens. J Occup Environ Hyg 2:143-54.
Ofner-Agostini, M., Gravel, D., McDonald, L. C., Lem, M., Sarwal, S., McGeer, A., Green, K.,
Vearncombe, M., Roth, V., Paton, S., Loeb, M. & Simor, A. (2006). Cluster of cases of
severe acute respiratory syndrome among Toronto healthcare workers after
implementation of infection control precautions: a case series. Infect Control Hosp
Epidemiol 27, 473-8.
Pippin, D. J., Verderame, R. A. & Weber, K. K. (1987). Efficacy of face masks in preventing
inhalation of airborne contaminants. J Oral Maxillofac Surg 45, 319-23.
Qian, Y., Willeke, K., Grinshpun, S. A., Donnelly, J. & Coffey, C. C. (1998). Performance of
N95 respirators: filtration efficiency for airborne microbial and inert particles. Am Ind
Hyg Assoc J 59, 128-32.
Rengasamy, A., Zhuang, Z. & Berryann, R. (2004). Respiratory protection against bioaerosols:
literature review and research needs. Am J Infect Control 32, 345-54.
31
Salgado, C. D., Farr, B. M., Hall, K. K. & Hayden, F. G. (2002). Influenza in the acute hospital
setting. Lancet Infect Dis 2, 145-55.
Schulman, J. L. (1967). Experimental transmission of influenza virus infection in mice. IV.

Relationship of transmissibility of different strains of virus and recovery of airborne
virus in the environment of infector mice. J Exp Med 125, 479-88.
Schulman, J. L. (1968). The use of an animal model to study transmission of influenza virus
infection. Am J Public Health Nations Health 58, 2092-6.
Seto, W. H., Tsang, D., Yung, R. W., Ching, T. Y., Ng, T. K., Ho, M., Ho, L. M. & Peiris, J. S.
(2003). Effectiveness of precautions against droplets and contact in prevention of
nosocomial transmission of severe acute respiratory syndrome (SARS). Lancet 361,
1519-20.
Tang, J.W. and Li, Y. (2007). Transmission of influenza A in human beings. Lancet Infect Dis
7: 758. (Author reply 761-3).
Tellier, R. (2006). Review of aerosol transmission of influenza A virus. Emerg Infect Dis 12,
1657-62.
Tellier R. (2007). Transmission of influenza A in human beings. Lancet Infect Dis 7: 759-60.
(Author reply 761-3).
Toner, E. (2006). Do public health and infection control measures prevent the spread of flu?
Biosecur Bioterror 4, 84-6.
Tuomi, T. (1985). Face seal leakage of half masks and surgical masks. Am Ind Hyg Assoc J 46,
308-12.
Weber, A., Willeke, K., Marchioni, R., Myojo, T., McKay, R., Donnelly, J. & Liebhaber, F.
(1993). Aerosol penetration and leakage characteristics of masks used in the health care
industry. Am J Infect Control 21, 167-73.
Webster, R. G., Guan, Y., Peiris, M., Walker, D., Krauss, S., Zhou, N. N., Govorkova, E. A.,
Ellis, T. M., Dyrting, K. C., Sit, T., Perez, D. R. & Shortridge, K. F. (2002).
Characterization of H5N1 influenza viruses that continue to circulate in geese in
southeastern China. J Virol 76, 118-26.
Wilde, J. A., McMillan, J. A., Serwint, J., Butta, J., O'Riordan, M. A. & Steinhoff, M. C.
(1999). Effectiveness of influenza vaccine in health care professionals: a randomized
trial. Jama 281, 908-13.
Yassi, A., Moore, D., Fitzgerald, J. M., Bigelow, P., Hon, C. Y. & Bryce, E. (2005). Research
gaps in protecting healthcare workers from SARS and other respiratory pathogens: an
interdisciplinary, multi-stakeholder, evidence-based approach. J Occup Environ Med
47, 41-50.
32
9 GLOSSARY
APF Assigned protection factor: The level of protection that 95% of properly trained
and supervised users of well maintained and correctly fitted RPE can expect to
achieve or exceed in real use situations. APF is conventionally represented by the
5th percentile of valid workplace or simulated workplace protection factor
measurements. The APF is the correct value to use when selecting RPE which is
capable of providing adequate levels of protection. See also NPF, PF and WPF.
FFP3 Filtering facepiece respirator achieving class 3 performance against airborne

particles.
Fit Factor A measure of the effectiveness of the faceseal of the respirator or surgical mask
against a wearer’s face.
Inert A measure of the inert particle concentration inside the respirator or surgical mask
particle compared to the particle concentration challenging the facepiece.
reduction
factor
Influenza A measure of the influenza virus titre in air sampled inside the surgical mask
plaque compared to the corresponding titre in the air challenging the facepiece.
reduction
factor
NPF Nominal protection factor. The level of protection achieved in laboratory

certification tests, assuming the maximum leakage permitted in the performance
requirement applied. Being measured under ideal laboratory conditions, this level
of protection is unlikely to be achieved in real-use situations, and should not be
used in the selection of equipment. See also APF and WPF.
WPF Workplace protection factor. The level of protection provided by an item of PPE
or ensemble, measured in real use conditions using appropriate methodology. With
a sufficient body of WPF data, the assigned protection factor (APF) is taken as the
fifth percentile of ranked WPF data. For technical or ethical reasons, it may be
impractical to measure WPF in real use situations. Simulation of realistic working
activity with a suitable tracer challenge agent is considered to be an acceptable
substitute for real WPF data.
33
34
35
Published by the Health and Safety Executive 04/08

Health and Safety
Executive

The UK is preparing for a potential influenza pandemic.

The main route of transmission of influenza is believed
to be via direct contact with large droplets. The relative
importance of aerosols in transmission is considered
to be minor, but it cannot be ruled-out. The current
UK Pandemic Influenza Infection Control Guidance
recommends that workers who are in close contact with
patients should wear surgical masks to reduce exposure to
large droplets. However, surgical masks are not intended
to provide protection against infectious aerosols. The
guidance recommends that procedures that are likely
to generate aerosols should be minimised, or where
unavoidable, workers should wear appropriate respiratory
protection. There is a common misperception amongst
workers and employers that surgical masks will protect
against aerosols. This study aims to evaluate the relative
levels of protection provided by both surgical masks and
respirators against aerosols.
This study focussed on the effectiveness of surgical masks

against a range of airborne particles. Using separate tests
to measure levels of inert particles and live aerosolised
influenza virus, our findings show that surgical masks
provide around a 6-fold reduction in exposure. Live
viruses could be detected in the air behind all surgical
masks tested. By contrast, properly fitted respirators could
provide at least a 100-fold reduction.
This report and the work it describes were funded by the

Health and Safety Executive (HSE). Its contents, including
any opinions and/or conclusions expressed, do not
necessarily reflect HSE policy.
RR619
www.hse.gov.uk

Health and Safety Executive 2008 Evaluating Protection Aforded by Surgical Masks

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Health and Safety Executive 2008 Evaluating Protection Aforded by Surgical Masks

Uploaded by

Copyright:

Available Formats

Health and Safety

Evaluating the protection afforded by surgical

Prepared by the Health and Safety Laboratory

Evaluating the protection afforded by surgical

Jonathan Gawn, Mike Clayton,

First published 2008

All rights reserved. No part of this publication may be

reproduced, stored in a retrieval system, or transmitted

in any form or by any means (electronic, mechanical,

photocopying, recording or otherwise) without the prior

written permission of the copyright owner.

Applications for reproduction should be made in writing to:

Licensing Division, Her Majesty’s Stationery Office,

St Clements House, 2-16 Colegate, Norwich NR3 1BQ

EXECUTIVE SUMMARY ................................................................................. VII

1.1 Background ....................................................................................................... 1

1.2 Respiratory Protective Equipment ..................................................................... 2

1.3 RPE in Healthcare and Pandemic Planning ...................................................... 3

2 MATERIALS AND METHODS.................................................................7

2.1 Inert Particle Tests On Test Subject .................................................................. 7

2.2 Respirators And Surgical Masks Employed In The Study .................................. 9

2.3 Influenza Bioaerosol Tests .............................................................................. 10

3.1 Inert Particle Exposure Tests On Test Subject ................................................ 13

3.2 Influenza Bioaerosol Tests .............................................................................. 18

4.1 Fitted Facepiece Respirators........................................................................... 20

4.2 Surgical Masks................................................................................................ 20

6 RECOMMENDATIONS FOR FUTURE WORK......................................25

6.1 Testing A Broader Range Of Surgical Masks And Respirators ........................ 25

6.2 Maximising The Protective Effect of Surgical Masks Against Influenza

6.3 Analysis Of Surgical Mask Protection Against A More Representative

6.5 Bioaerosol Challenge To FFP Respirators ...................................................... 26

Feb May 2006) ................................................................................................ 27

May 2006) ....................................................................................................... 28

The relative contribution of aerosol transmission in natural influenza transmission is thought to

In theory, influenza can be transmitted via three major routes

• Large droplets (>5µm diameter)– these settle rapidly on surfaces

• Manual inoculation/direct contact or contact with secretions

1.2 RESPIRATORY PROTECTIVE EQUIPMENT

• Use of personal protective equipment (PPE)

• “CE” marked to the European PPE Directive

• Selected, used and maintained by properly trained people

• Correctly maintained, examined and tested

Table 1. Efficiency requirement for filtering facepieces and their assigned

1.3 RPE IN HEALTHCARE AND PANDEMIC PLANNING

2.1 INERT PARTICLE TESTS ON TEST SUBJECT

2.1.2 Inert aerosol generation

2.1.3 Performance measurement

2.1.3.1 TSI Portacount

2.1.3.2 Test procedure

The following tests and measurements were undertaken:

A schematic diagram of the test arrangement is shown in Figure 2.1.

Figure 2.1. Schematic diagram of the inert aerosol test arrangement

2.2 RESPIRATORS AND SURGICAL MASKS EMPLOYED IN THE STUDY

Figure 2.2. A diagrammatic representation of those surgical mask designs

2.3 INFLUENZA BIOAEROSOL TESTS

2.3.1 Virus Culture and Processing

2.3.2 Inert aerosol sampling

2.3.3 Live influenza bioaerosol sampling

2.3.4 Sample processing

Figure 2.3. Diagrammatic representation of the influenza bioaerosol testing rig

3.1 INERT PARTICLE EXPOSURE TESTS ON TEST SUBJECT

3.1.1 Calculation of the reduction factor and fit factors

Particle concentration outside the facepiece (Co)